The lower image is 23% of the area in the image above it. Taken with the Leica X1 this demonstrates that the lens/sensor combination has great potential as a KAP camera. The image quality has clearly been helped by the strong light in this example but, none the less, this is evidence of a camera with an above average performance for its weight. This is an AutoKAP image. This blog takes a look at why the X1 comes out as a good candidate for KAP, and what needs to be done to get it to work as a directed camera.
Lightweight camera with good resolution: how hard can it be? Aerial photography by kite has some very specific requirements to be met by the camera: the 3 biggest problems the camera needs to cope with are:
1.Weight. This is obvious, a heavy thing won’t fly as well as a light one: look at a birds bones..they are hollow!
2.Resolution. This is tricky to define but taking pictures at 30-60m remote from the subject is not what most cameras are designed to do. Air survey cameras deal with the resolution problem by using very large lens arrays and big sensors to get the most from the vertical vantage point at 600m or more above ground.
This is fine if you have a powerful aircraft ( this example is an F49 as fitted in the English Electric Canberra, fine indeed!) to lift a large camera, I do not! My aerial platform of choice generates between 2 and 10kg of lift. Resolution is dependant on lens quality, sensor performance and optical design but the KAP camera needs to achieve this in the lightest package possible. (for more on the Canberra air cameras see here)
3.Remote controllability. Although it is fun to get pictures from on high by automatic camera control (AutoKAP) I feel the effort in getting the camera in the right place at the right time is best spent with a camera I can point at the subject with confidence: this is a personal preference and I have the greatest respect for those who practice KAP control by a honed skill set requiring judgement and timing to get the shot they want but I need a camera I can use remotely by seeing what the camera sees to get the shot I want. Ideally a camera with an ‘AV’ out socket of some kind will enable a video down link, if this is not present then a ‘pilot’ mini cam will do the job.
James Gentle has a good page on this:http://www.gentles.info/KAP/rig04/cameras.ht
the discussion continues at:
The perfect mix of minimum weight, a high pixel count, sensor size and pin sharp lens is an elusive thing. For many the Canon G series hits the spot, the Micro 4/3rds cameras are good (the Olympus EP1 in particular is a very flexible camera indeed). Here are the runners and riders in the KAP sub DSLR contest:
|Camera||Weight||Key feature||Sensor size||Sensor mm2 per gm|
|Sigma DP1/DP2||278gm||14Mp||20.7 x 13.8 mm||1.027|
|Ricoh ‘Calipo’ GX200||238gm||12Mp||7.60 x 5.70 mm||0.18|
|Ricoh GXR ‘S10’ or ’P10’||397gm with 50mm ‘A12’ lens option||12Mp. Interchangeable lens/sensor packs available. AV out connection||23.6 x 15.7mm||0.93|
|Canon G11||335gm||10Mp programmable||7.60 x 5.70 mm||0.12|
|Panasonic Lumix DMC-LX3||228gm||Leica Wide Angle lens. Image stabilisation||8.17 x 6.13 mm||0.21|
|Olympus EP1||406gm*||Interchangeable lenses. Micro 4/3rds format. ‘On board’ image stabilisation. 2/3rds full frame CCD.||18 x 13.5mm||0.598|
|Panasonic DMC GF1||485gm*||Interchangeable lenses. Micro 4/3rds format. ‘On lens’ stabilisation||0.50|
|SonyNex 5||317gm*||Interchangeable lenses. Sony ‘E’ mount 4/3rds format. 14Mp||23.4 x 15.6 mm||1.12|
|Leica X1||286gm||12Mp. Fixed 35mm equivalent focal length lens.||23.5 x 15.7 mm||1.29|
*includes lightweight wide angle ‘pancake’ lens.
‘Mirco4/3rds’ is a specific term for the ‘standard’ miniature interchangeable lens mount whereas ‘4/3rds’ is used to describe the sensor size ( i.e.: sensors of approx 4/3rds of an inch diagonal) The Micro 4/3rds lenses available are designed to be lightweight and compact to match the small camera body. Clearly this is a benefit for KAP, smaller and lighter cameras with better resolution are exactly what’s needed!
So far only Olympus and Panasonic have adopted this standard, Sony have decided to go their own way and Leica already have their own ‘M’ system. To date I have tried the Olympus ‘Pancake’ wide angle and the collapsible 14-42mm; they are clearly compromise designs and have aberration and light loss problems respectively: but they are very light compared to the DSLR equivalents. The Micro4/3rds lens mounting system has a lot going for it and new lenses are on their way, for KAP the weight saving of using a camera with out a prism and mirror is obvious.
Leica X1 has best ‘weight per pixel’ score. So gram for gram the X1 has more sensor than any of its rivals, for every gram of weight the camera carries a massive 1.29 mm2 of sensor. This is, of course, not the only consideration in choosing a KAP camera, the X1 with its fixed ’35mm equivalent’ lens will be limiting for some and the quality of the sensor, its sensitivity and processing speed are also critical factors. It is not a flexible camera in the way the Micro4/3rds designs are either.
The bigger the sensor the better the resolution; in crude terms this can be likened to the negative size from the days of film: bigger is better, more light is collected and the ‘grain’ size will be smaller in proportion to the image area.
Opinion is divided on camera weight in KAP circles: some say the weight is not a problem, just match the kite to the load and the wind, maintaining the balance by bigger kites at lower wind speeds.
My experience doesn’t bear this out. In light winds the airflow can fluctuate from a happy 5 or 6 mph to a sudden gut wrenching zero: there is no getting away from the awful fact that when the wind drops so does your camera, it is true there is a glide effect with some kites which can save the day and it is also true that the Flowform can act as a parachute too BUT the bigger the load on the kite the faster the rig will fall! Careful design of kite can mitigate some of the worst effect of loss of lift. At KAPiNed10 Ralf Beutnagel impressed me with his concern in designing out the collapse characteristic of Flowforms in engineering his superb ‘KAPfoil’ http://www.dopero.de/Eingang/Meine_Drachenseite/Ralfs_KAP-Foil/ralfs_kap-foil.html
So my reaction to seeing the rig drop 20-30m in seconds when a light wind lets go is to treat weight as an enemy…true I use an overweight line (250lb /113kg solid core woven Dacron polyester…currently unobtainable!) but I cannot risk line failure so the safety margin there brings a weight penalty.
In the war on weight I am experimenting with a new Climax ‘Powerline Extreme’ line from Emkay Kites
Rated at a remarkable 250DaN (1 daNewton = 10 Newtons = 1Kg so approx 250kg) from Germany which is very light but difficult to handle: it’s like steel wire under load- impossible to grip and very sharp. It’s so fine I can get half a kilometre on a 10″ reel:
The flying angle of the ff30 is quite low ( 30 to 40 deg from the horizon) the super light Climax line improves this by about 5 degrees in light wind (6-7 mph). The low angle is something of an advantage in KAP, provided there is space for the line, I can get over buildings downwind of the launch site reasonably comfortably. Unlike the Dacron, the Climax Powerline Extreme line gets ‘jangly’ and starts thrumming just as the ff30 starts to stiffen up (at around 8mph wind speed) and this is imparts vibration into the rig.
Remote control of shutter. So back to the camera. The X1 is problematic as a KAP camera. Leica, in their wisdom, have failed to supply the camera with a live video output. It doesn’t have an IR control option either so a servo is needed for the shutter. No problem there; a bit of quality shed time gets you there. It’s not elegant I know but Leicas have had all manner of odd stuff plonked onto them from wire hoop frame guides to capstan range finders so I think my ally bashing is in keeping with the ‘M’ heritage!. Hitec mini servo, 3 snap rivets, a cork spacer and 3 bits of ‘L’ section get the job done for less than 20g:
The first thing I noticed when I sifted though the early results was the high number of poor exposures (mostly motion blur) with the camera set to ‘Auto’: the exif (http://en.wikipedia.org/wiki/Exchangeable_image_file_format) data shows it wants to work at slow ISO speeds (usually 100) so clearly it needs to have the ISO upped and likewise the shutter speed. Experiment shows ISO 400 and shutter at 250th/s in strong light works well. Unusually for a ‘compact’ this camera does not work well using the ‘default’ settings!
So all that’s needed is to get the preview video link and we have the perfect KAP camera? Maybe: the weight is piling on now and the once glorious 286g is now 300g and climbing.
Using a cheap mini video cam as a preview. Fitting a ‘pilot’ mini-video camera requires solving 3 problems:
1.Voltage conditioning. My rig is powered by a 5 cell NiMh pack from Ripmax rated at 6v/2000mA (it often runs at 6.2v according to James Gentles AltiOSD board). The rig servos are rated from 4.8 to 6V. The miniature video cameras are quoted as running from 9 to 12 V but operate over a more restrictive range of 7- 8.5V as sold. Getting the required voltage and current measured and regulated is crucial. The mini video camera we bought uses an in line regulator to 5 V then another on the camera board with a 3.3V output. Modified to use a 3.3V regulator with the infra-red L.E.D’s removed, it draws 130mA and cuts out as the voltage drops below 5.8v.
2. Safe framing. Ideally I want to see the same view on the ground as the camera sees in the air. Getting the pinhole view to match the X1 view may require a mask to be placed in front of the pilot camera.
3. Alignment and fixing. The pilot video needs to be demountable and re-mountable such that its framing is consistent from shoot to shoot. Any fixing needs to be of minimal weight as well as robust. Most of my kit seems to get damaged being lugged about in the camera case rather than flying! My rig needs to be able to take a selection of cameras ( ideally 2 but I’m working with 3 at the moment: Leica DeLux 4, Olympus EP1 and the X1) according to the lift available, so the pilot camera should mount and de-mount without fuss and survive life in transit.
Fresh out of the Hollins workshop the bolt-on/bolt-off mount is of 0.07mm aluminum sheet which is screwed onto the rig using the camera tripod mounting (Jesus) nut. The assembly weighs in at 17gm bringing the X1 payload up to 317gm.
Mini video ‘spy’ cameras are cheap on ebay, they come in a variety of set-ups; some have focusable lenses, some pin hole lenses, some are rated at 9V others 12V so understanding what is in the box is the 1st step in getting the preview working. The supplied voltage regulator (the lump in the cable) was found to be superflous as its function seemed to be simply to uprate the camera for 9V working. Stripped to ‘bare wires’ the camera was found to be working at 5.8 to 7 Volts, a power supply cable with a Futaba pattern connector straight to the rig battery (6.2V) was fitted. The video feed cable was also replaced with a screened one to reduce interference. The camera mounted LEDs are stripped (we won’t be flying in the dark!) to reduce weight and current uptake.
Two cameras were aquired for the task, one in a heavy metal case with 6 LED’s and a ‘fish eye’ lens and the other with a plastic shell with a ‘pin hole’ type lens and 4 LED’s. Stripping down the cameras revealed a number of common parts alowing the lighter (plastic case) to be joined with the fish eye lens from the heavier camera giving a field of view closest to X1 for the minimum weight.
The rig set up for the first directed camera flight: power to the mini cam is taken direct from the battery to the left and the video cable is attached to the transmitter on the right. The alignment of the minicam is fixed by location through the tripod fixing mount. An alternate mounting on the shutter servo mount will be set up once the depth position and vertical paralax of the minicam is better understood by experiment with the bolt on-bolt off fixing. As set up the width of the view is a good match to the X1 FoV and the height is slightly oversize.
So without further ado the rig flew today and the 1st lesson of pilot video was learned: the camera exposure control was jammed between 2 settings and I spent a happy half hour thinking I was getting really good range and frame test shots when in fact I was shooting blanks! Yep each shot which was perfectly composed on the ground was stored on the SD card as a neat white square….The kite is a 3m rokkaku by Skybums (more on Skybums here ) comissioned by Clive Hollins which works beautifully when nothing else will lift the rig, in this case in 3mph wind speed.
The next episode will reveal the first shots, but now I have to wait for the weather! Which will give me time to tidy those loose wires and fit the minicam to the shutter servo mount.